“Cucumber is Dying”, What did we learn? Part 1: The Facts

The article title is not clickbait. It is the headline of the Software Testing Weekly #166 (on 2023–05–01).

“Back in February, Matt Wynne — the last of Cucumber’s co-founders who actively worked on the project — got fired. He shared about it in this blog post.

That raised questions about the project’s future. Someone even claimed that Cucumber is dead which sparked a big discussion on Reddit.

So what’s next? Without a bigger sponsor behind the project, the chances for its growth and maintenance are lower.

On the other hand, thousands of people and companies still use it. So now it’s up to the community.

But I’d think twice before picking Cucumber again.”

While the official announcement (like previous failed Phantom.js and Protractor) is not out yet, it is fair to say “Cucumber is dying”. I will elaborate on Matt’s blog post shortly.

Firstly, as a programmer who has worked on open-source projects, I feel pity for the creators of Cucumber. While I always think “Gherkin is bad for end-to-end test automation” (see my article: “Why Gherkin (Cucumber, SpecFlow,…) Always Failed with UI Test Automation?”, written in Jan 2021 and my book ‘Practical Web Test Automation with Selenium WebDriver’ 2nd edition in 2013), I have big respect for Cucumber creators, especially, Aslak Hellesøy, whose DamageControl CI server project (not officially released) inspired me to create BuildWise CT server using the wonderful language: Ruby.

Please note, my opinion against Cucumber or Gherkin is limited to the scope of end-to-end test automation, which I am familiar with.

“So really, what is Cucumber? As a test tool it sucks. There far better automated test tools” (source) — Aslak Hellesøy, creator of Cucumber

In this article, I will answer some questions some might have. Before that, to get a reality check about the current situation of Cucumber, I will quote some words from its co-founder Matt Wynee’s post.

The whole original Cucumber team was gone: “As the last remaining employee lucky enough to be paid to work on the open source project full-time”

It is about to end: “I figured we might be in the end-game, but I hadn’t expected it to come around so suddenly.”

Sacked with notice: “I was actually on vacation, with no internet, while all this happened last week, so I was not prepared. I’ve lost access to my [email protected] email address”

“The future of the project is uncertain right now.”

The failure of Cucumber might be a surprise to some, but not to me at all. I predicted this 10 years ago, in my book ‘Practical Web Test Automation with Selenium WebDriver’ 2nd edition in 2013. For why, check out my article, Why Do Most UI Test Automation Fail? (Part 2: Wrong Choice of Test Syntax Framework).

By the way, I do know quite a bit about Cucumber, as I added Cucumber support to TestWise: my functional testing IDE and BuildWise: my Continous Testing Server. In other words, my prediction is based on open-minded perspective.

Considering the following facts:

• SmartBear has money, resources (human and integration with its existing test automation suites)
• Cucumber Team (residing in SmartBear) has the best knowledge of Cucumber BDD, and now has access to more resources after the acquisition.
• Cucumber/BDD is well-known in the software industry. With SmartBear’s sales channels in the test automation space, (if it works), commercializing it shall be easy. At least, SmartBear management thought so before the acquisition.

Yet, it failed. If SmartBear had consulted with me before 2019, or read my book, it would save them quite a lot of money. The reason for my correct prediction is simple: I have acquired a deep understanding of End-to-End Test Automation and Continuous Testing via doing real test automation since 2006, on a daily basis.

In Part 2, I will answer some questions you might have.

--

This article was originally published on my Medium blog on 2023-05-05.